TiMeDDx--a multi-phase anchor-based diagnostic decision-support model

Main Clinical Manifestation (MCM)-oriented diagnosis starts with a chief problem and reasons about possible diagnoses that can be manifested in that way. The reasoning process often starts by considering abstract diagnosis groups (e.g., infectious vs. non-infectious diarrhea) and refines them. Most existing diagnostic decision-support systems (DSSs) are not specially tailored toward assisting non-expert physicians in the proper and efficient investigation workup of MCM-oriented diagnosis. We developed a prototype diagnostic decision-support model called TiMeDDx that is MCM-oriented and follows the hypothetico-deductive clinical reasoning process of differential diagnosis. The model guides users in a phase-by-phase manner regarding abstract diagnosis groups and diagnoses that should be considered and appropriate data that should be collected during the clinical investigation process. TiMeDDx's knowledge base contains, when possible, knowledge derived from MCM-oriented evidence-based sources. We explain the knowledge model and diagnostic algorithms (Bayesian and heuristic) of TiMeDDx, using the clinical problem of diarrhea as a case study, and contrast TiMeDDx with models of existing diagnostic DSSs.